For e-commerce · catalogue-scale AEO

Be one of the three products it recommends.

Shopping assistants don’t return a results page — they return a shortlist. If your price, stock and returns policy aren’t machine-readable, you are not on it, however good the product is.

The engagement

Audit, fix at template level, track orders.

One change, forty thousand products.

01
Feed audit
Every product template checked the way a shopping assistant reads it: price, stock, variants, returns, reviews and identifiers.
02
Fix the catalogue
Offer schema completed at scale, variants given their own URLs, policies moved out of PDFs and into readable pages.
03
Win the buying questions
Comparison, sizing and “which one should I get” pages built for the prompts your category actually gets.
04
Track what moved
A monthly report: mention share per collection and where your products now appear across the assistants.
What we fix

Six gaps that keep good products off the shortlist.

All of them are catalogue-wide, and all of them are fixable at template level.

Complete Offer data

Price, currency, availability, shipping and returns published as data — the fields assistants compare on before anything else.

Variants with their own URLs

Size and colour hidden behind query parameters are invisible. Each buyable variant gets a crawlable, quotable page.

Live from your feed

Your product feed keeps price, stock and availability current automatically — so an assistant never recommends a sold-out item or quotes yesterday’s price.

Product identity

GTIN, MPN and brand set consistently so your listing is matched to the right product across every source.

Comparison content

The “X vs Y” and “best under £80” pages that assistants quote when a buyer is deciding.

Shortlist tracking

See how often your products get named versus rivals — per collection, across ChatGPT, Gemini, Claude and Perplexity — in a monthly report.

Before / after

What changes in the markup.

A typical product template on a JavaScript-rendered store.

Typical product page
{"@type":"Product",
 "name":"Merino Base Layer"}

// price rendered by JS
// no availability
// variants behind ?colour=
After
{"@type":"Product",
 "name":"Merino Base Layer",
 "gtin13":"5060123456789",
 "offers":{"@type":"Offer",
   "price":"74.00","priceCurrency":"GBP",
   "availability":"InStock",
   "hasMerchantReturnPolicy":{"returnDays":60}},
 "aggregateRating":{"ratingValue":"4.7",
   "reviewCount":"812"}}
Old funnel vs. new

Shopping stopped being a results page.

The buyer sees three names, not ten links.

 
Search shopping
Assistant shopping
Buyer starts with
A search box
A question
You compete on
Ranking position
Being one of three names
Product data needed
Title, image, price
Price, stock, returns, GTIN, reviews, variants
Content that wins
Category pages
Comparison and “which should I buy” answers
Attribution
Referral traffic
Assistant referrals and direct brand searches
Speed of change
Weeks
Days — answers refresh continuously
Platforms

Works with what you already run.

Shopify
WooCommerce
Magento
BigCommerce
Shopware
Custom / headless
From merchants

What moved.

Adding availability and returns to our Offer schema was a two-day job. Six weeks later we were in the ChatGPT shortlist for our two biggest category questions.
Sofia Herrmann
Head of Digital, Brackenfield Outdoor
The variant URLs alone were worth it. Assistants can now recommend a specific size in stock instead of skipping us entirely.
Callum Reece
E-commerce Manager, Vantry
Questions

Before you commit.

No checkout changes. Most work is schema and template-level markup, delivered as a patch your developer or agency can review before it ships.

Fixes are made at template level, so one change covers the whole catalogue. Feed validation then runs continuously against every product.

We track mention share per collection and exactly where your products appear across the assistants, month over month. We don’t claim to attribute individual orders — we show the shortlist presence that leads to them.

No — it complements them. Assistants read your site directly as well as merchant feeds, and most stores fail on the site side.

Feed fixes are picked up within days. Comparison content typically shifts answers in four to six weeks.

Start with one product page.

Setup is fast. Cancel anytime.

Monthly plans from $149. Cancel anytime.